jojochuang commented on PR #11054:
URL: https://github.com/apache/ozone/pull/11054#issuecomment-5502358988

   ## Config behavior summary (this PR vs HDDS-15072)
   
   `decommission.ec.reconstruction.enabled` and `load.factor` are **defined but 
not read** by SCM in this PR — behavior is unchanged regardless of their values 
until [HDDS-15072](https://issues.apache.org/jira/browse/HDDS-15072).
   
   `reconstruction.global.limit` is **enforced** in 
`sendThrottledReconstructionCommand()` (missing-index EC reconstruction path 
only). It does **not** affect decommission 1-1 replication 
(`processDecommissioningIndexes` → `createReplicateCommand`).
   
   | Scenario | EC decommission | EC missing-index reconstruction |
   |----------|-----------------|--------------------------------|
   | Defaults (`enabled=false`, `limit=0`) | Same as before | Same as before 
(no global cap; inflight count tracked for observability) |
   | `enabled=true`, `limit=0` | Same as before (`enabled` ignored) | Same as 
before |
   | `enabled=false`, `limit>0` | Same as before | **New:** cluster-wide 
reconstruction cap enforced |
   | `enabled=true`, `limit>0` | Same as before (`enabled` ignored) | **New:** 
cluster-wide reconstruction cap enforced |
   
   When the global cap is hit, reconstruction is deferred 
(`CommandTargetOverloadedException`); the container is retried on the next 
replication monitor cycle. Decommission continues via 1-1 replicate and is not 
subject to this cap.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to